Drug therapy for many problems (cancer, AIDS, infectious disease, transplants, cardiovascular disease, stroke, etc.) is limited too often by drug toxicity. Therapy can be optimized for an individual patient by describing quantitatively the relationships between dose, drug exposure (serum levels and/or area under the serum level curve, AUQ and effect, such as bacterial, viral and cancel cell growth and killing, as described by Zhi (and many other) models, as well as those between drug exposure and toxicity. Development of dose-exposureeffect relationships allows rational selection of target therapeutic goals such as serum drug levels or profiles, or drug effects, and thus permits dosing to proceed optimally, and with optimal precision in their achievement. The barrier to such optimized drug therapy has been the lack of dose-concentration-effect relationships developed from populations of patients studied in their relevant real-life clinical care situations. Population pharmacokinetic and dynamic modeling is the tool to remove this barrier.